Prev: optimize, inline, oh my (questions)
Next: ITA may be sold
From: His kennyness on 25 Apr 2010 10:08
Captain Obvious wrote:
> Hk> Change yer frickin name. This behavior is utterly non-obvious
>
> I'm not saying it is.

Perhaps I misconstrued "well-known idiom"? If all you meant is, yeah,
that's f*cked up, but hard-core loopers know about it" then (a) we agree
and (b) I do not understand how that is an objection to a loop rant.

What is more interesting is that Mr. Margolin reports dotimes does the
same, and I have never run into that there, tho I eschewed loop for the
first ten years of my Lisp careeer. Implications unclear.


hk
From: Tim Bradshaw on 25 Apr 2010 10:38
On 2010-04-25 15:08:59 +0100, His kennyness said:

> Perhaps I misconstrued "well-known idiom"? If all you meant is, yeah,
> that's f*cked up, but hard-core loopers know about it" then (a) we
> agree and (b) I do not understand how that is an objection to a loop
> rant.

I don't understand this whole thread. Why would anyone assume either
that each iteration creates a new binding, or that the same binding is
reused? Neither seems particularly better to me in the abstract, and I
can imagine a language specification which just left this undefined[*].
So, obviously, anyone who wanted to capture such bindings would wrap
LET around their code. I mean, seriously, who would not do that?

--tim

[*] Such as, for instance, the Common Lisp specification (see DOTIMES,
DOLIST).

From: Tamas K Papp on 25 Apr 2010 10:43
On Sun, 25 Apr 2010 15:38:48 +0100, Tim Bradshaw wrote:

> On 2010-04-25 15:08:59 +0100, His kennyness said:
>
>> Perhaps I misconstrued "well-known idiom"? If all you meant is, yeah,
>> that's f*cked up, but hard-core loopers know about it" then (a) we
>> agree and (b) I do not understand how that is an objection to a loop
>> rant.
>
> I don't understand this whole thread. Why would anyone assume either

The purpose of this whole thread is to vent frustration that arose
from making an unwarranted assumption.

> that each iteration creates a new binding, or that the same binding is
> reused? Neither seems particularly better to me in the abstract, and I
> can imagine a language specification which just left this undefined[*].
> So, obviously, anyone who wanted to capture such bindings would wrap
> LET around their code. I mean, seriously, who would not do that?

Precisely.

Best,

Tamas
From: Steven E. Harris on 25 Apr 2010 11:27
Tim Bradshaw <tfb(a)tfeb.org> writes:

> Why would anyone assume either that each iteration creates a new
> binding, or that the same binding is reused?

I'm reminded of this thread from October 2002:

Midfunction Recursion
http://groups.google.com/group/comp.lang.lisp/browse_frm/thread/422226bb18e56cad/19fae84599dd027a

I always remember -- fondly -- the threads involving direct
correspondence with Erik Naggum.

--
Steven E. Harris
From: Pascal Costanza on 25 Apr 2010 12:23
On 25/04/2010 16:38, Tim Bradshaw wrote:
> On 2010-04-25 15:08:59 +0100, His kennyness said:
>
>> Perhaps I misconstrued "well-known idiom"? If all you meant is, yeah,
>> that's f*cked up, but hard-core loopers know about it" then (a) we
>> agree and (b) I do not understand how that is an objection to a loop
>> rant.
>
> I don't understand this whole thread. Why would anyone assume either
> that each iteration creates a new binding, or that the same binding is
> reused? Neither seems particularly better to me in the abstract, and I
> can imagine a language specification which just left this undefined[*].
> So, obviously, anyone who wanted to capture such bindings would wrap LET
> around their code. I mean, seriously, who would not do that?

Each iteration having a new binding can be an advantage in the abstract,
since updating bindings creates (unnecessary) dependencies between the
loop branches, which could be an issue when attempting to automatically
parallelize loops.


Pascal

--
My website: http://p-cos.net
Common Lisp Document Repository: http://cdr.eurolisp.org
Closer to MOP & ContextL: http://common-lisp.net/project/closer/
First  |  Prev  |  Next  |  Last
Pages: 1 2 3 4 5
Prev: optimize, inline, oh my (questions)
Next: ITA may be sold